Amos 21

For PSI-II, confirmatory factor analysis was conducted with IBM AMOS 21, constraining the instruments into three dimensions stated in the original instrument. The indices recommended by Schreiber et al. [29 (link)] to indicate an adequate model fit were goodness-of-fit (GFI), comparative fit index (CFI), root mean square error of approximation (RMSEA), and they yielded acceptable scores (GFI = 0.95, CFI = 0.97, TLI = 0.92, and RMSEA = 0.056).
For the Tromsø social intelligence scale, confirmatory factor analysis was conducted with IBM AMOS 21, constraining the instruments into three dimensions stated in the original instrument. The indices recommended by Schreiber et al. [29 (link)] to indicate an adequate model fit were goodness-of-fit (GFI), comparative fit index (CFI), and root mean square error of approximation (RMSEA), and they yielded acceptable scores (GFI = 0.97, CFI = 0.99, TLI = 0.95, and RMSEA = 0.044).
For the IAT, a confirmatory factor analysis was conducted with IBM AMOS 21, constraining the instruments into six dimensions stated by the authors. The indices recommended by Schreiber et al. [29 (link)] to indicate an adequate model fit were goodness-of-fit (GFI), comparative fit index (CFI), and root mean square error of approximation (RMSEA), and they yielded acceptable scores (GFI = 0.93, CFI = 0.95, TLI = 0.93, and RMSEA = 0.067).

The data were analyzed using AMOS 21.0 and SPSS 26.0 statistical programs. First, descriptive statistical analyses and correlational analyses were performed to identify the general trends and characteristics of the data (such as mean and standard deviation). Next, factor analyses were performed to verify the construct validity of the surveys, and reliability analysis was performed for each factor (Cronbach’s α). Furthermore, to measure changes in the participants’ healthy lifestyle, a paired t-test was performed. To test whether autonomy in exercise participation and self-control had a significant impact on posttest healthy lifestyle (T2) even after controlling for the impact of pretest healthy lifestyle (T1) on posttest healthy lifestyle, hierarchical regression analyses were performed.